Available online at www.sciencedirect.

com

Composites: Part A 39 (2008) 875886

www.elsevier.com/locate/compositesa

Renewable resource based biocomposites from natural ber

and polyhydroxybutyrate-co-valerate (PHBV) bioplastic
Sanjeev Singh a, Amar K. Mohanty a,*, Tomohiko Sugie b, Yoshihiro Takai b,
Hiroyuki Hamada b
a
School of Packaging, Michigan State University, 130 Packaging Building, East Lansing, MI 48824, USA
b
Division of Advanced Fibro-Science, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto, Japan

Received 28 July 2007; received in revised form 28 December 2007; accepted 10 January 2008

Abstract

Renewable resource based green biocomposites were prepared using a bacterial polyester i.e., poly(hydroxybutyrate-co-valerate)
(PHBV) and natural bamboo ber. Fabrication of the biocomposites was carried out by injection molding following extrusion com-
pounding of PHBV and bamboo ber with 30 or 40 wt.% ber. The mechanical, thermo-mechanical and morphological properties of
the biocomposites were evaluated. Little variation in the thermo-mechanical and impact properties was observed when the ber content
was varied. The tensile modulus of biocomposites at 40 wt.% ber improved by 175% as compared to that of neat PHBV. The theoretical
tensile modulus of the biocomposites was calculated using Christensens equations and compared with the experimental results. It was
found to be in near approximation to the experimental data. The storage modulus was aected slightly by the variation of ber content
from 30 to 40 wt.% in biocomposites. The heat deection temperature of PHBV increased by 9 C at 40 wt.% of ber reinforcement.
Morphological aspects and thermal stability were studied using scanning electron microscopy and thermo-gravimetric analysis, respec-
tively. In addition, a comparative analysis of bamboo berPHBV with wood berPHBV biocomposites was performed. Statistical
analysis of both biocomposites was carried out by performing a two-way ANOVA on their tensile and exural moduli in order to eval-
uate the eect of ber type and content in the PHBV matrix.
2008 Elsevier Ltd. All rights reserved.

Keywords: A. Wood; A. Fiber; A. Thermoplastic resin; B. Mechanical properties

1. Introduction for these materials. Contemporary W/NF PC are based

In the established market of plastic composites there has
been a dynamic emergence of wood and natural ber plas-
tic composite (W/NF PC). Not only do they have a proven
market in outdoor and indoor applications like decking,
railing, fencing, furniture, etc., they also hold a large share
of the European market in automotive applications [1].
Most of thermoplastics used in W/NF PC are recyclable
and are being looked upon as a protable opportunity
*
Corresponding author. Present address: Department of Plant Agricul-
ture & School of Engineering Crop Science Building, University of
Guelph, ON, Canada N1G 2W1. Tel.: +1 517 355 3603; fax: +1 517 353
8999.
E-mail addresses: mohantya@msu.edu, mohanty@uoguelph.ca (A.K.
Mohanty).
on conventional thermoplastics and naturally available cel-
lulose ber, with the ber content varying from 30 to
70 wt.%. Polypropylene has the major market share of
W/NF PC followed by polyethylene, polyvinylchloride
and other resins. Various kinds of natural and wood ber
have been used to achieve good mechanical properties of
W/NF PC. A signicant amount of research is focused
on improving the compatibility between the hydrophilic
natural ber and the hydrophobic thermoplastics [25].
Nano-reinforcement using inorganic material like clay in
wood-PP composites has also been reported elsewhere [6].
The major limitation these composites possess is their lim-
ited dimensional stability when exposed to moisture [7].
There are environmental concerns due to the accumula-
tion of plastic wastes over last few decades; therefore

1359-835X/$ - see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2008.01.004
876 S. Singh et al. / Composites: Part A 39 (2008) 875886
biodegradable green plastics are being looked upon as possi-
ble substitutes to petroleum-based plastics. Polyhydroxyalk-
anoates (PHAs) are a family of natural, biodegradable
polyesters which are intracellularly synthesized as an energy
reserve by various bacteria. PHAs dier from petroleum-
based synthetic polymers such as polypropylene and poly-
ethylene in regard to their renewable resource based origin
and biodegradability. Polyhydroxybutyrate-co-valerate
(PHBV) a member of the family of PHAs has gained a lot
of attention as a green material. It is a brittle and crystalline
material with short pendent side-groups on its backbone.
This thermoplastic polyester is composed of hydroxybuty-
rate (HB) units with between 0% and 24% of hydroxyvalerate
(HV) units appearing randomly throughout the polymer
chain [8]. The similarity of some of its mechanical properties
to the polyolens indicates that it may be a substitute for
polyolens. The biodegradability and biocompatibility of
PHBV have given it a remarkable breakthrough in biomed-
ical applications [9]. Its composites with natural ber have
been studied extensively [1012]. Wood and other natural
bers are generally used to reinforce thermoplastics due to
their low cost, abundant availability, high performance
and low density [12]. Among the existing WPC products,
maple wood ber is the most widely used commercial ber;
whereas among the naturally available bers in Asia, bam-
boo ber has the most abundant applications [13,14]. There-
fore, these two bers, obtained from two dierent categories
of lignocellulose-based sources are the most used bers in the
commercial sector.
The prime focus of this study was to fabricate bamboo
ber based PHBV biocomposites. Fundamental aspects
of these biocomposites were measured using dierential
scanning calorimetry (DSC), thermo-gravimetric analysis
(TGA), Fourier transform infrared spectroscopy (FTIR),
and their morphology was also explored. Furthermore,
the experimental results of tensile modulus were compared
with the existing theoretical model. Comparative analysis
of the processed bamboo ber biocomposites with the ear-
lier reported wood berPHBV biocomposites [15] has also
been discussed; along with performing a two-way ANOVA
on their mechanical properties.
2. Experimental section
2.1. Materials
Polyhydroxybutyrate-co-valerate (PHBV) under the
trade name Biopol (Zeneca Bio Products) was obtained
from Biomer, Germany. Madake bamboo ber with lengths
of 500 lm and diameters between 10 and 100 lm was sup-
plied by the Kyoto Institute of Technology, Japan.
2.1.1. Fabrication of composites employing micro-
compounding technique
In order to avoid any moisture induced degradation dur-
ing processing, the bamboo ber and PHBV were dried at a
110 C and 80 C under vacuum for 2 and 4 h, respectively.
Table 1
Processing parameters of PHBV and ber compositions
PHBV Fiber Processing temperature Retention time
(wt.%) (wt.%) (C) (min)
100 0 160 3
70 30 160 10
60 40 160 10
The dried bers and PHBV were kept in air-tight plastic
bags until the time of processing. The composites were pre-
pared using a micro-extruder (DSM Research, Netherlands)
with a barrel volume of 15 cc. The extruder was equipped
with co-rotating twin-screws having a length of 150 mm
and L/D (length to diameter ratio) of 18. The micro-extru-
der had three consecutive heating zones and the processing
temperatures in these three heating zones (top, center and
bottom) were 165, 160 and 155 C. The processing temper-
ature was kept between 155 and 165 C in order to achieve
low melt ow and to avoid thermal degradation of PHBV at
higher temperature. The screw rotation speed of the micro-
extruder was xed at 100 rpm. PHBV pallets and the ber
were physically mixed as per the listed ratios in Table 1
and fed to the micro-extruder through the feeding zone
(top). After the xed processing time in the micro-extruder
(mentioned in Table 1), the molten mix was transferred to a
preheated mini-injection molding machine for specimen
fabrication. The optimized injection pressure to extricate
the melt from the cylinder to the mold located in the mini-
injection molding machine was 551.5 KPa at the tempera-
ture of 155 C. Finally, the specimen was removed from
the mold and characterized. The dierent mix ratios of
PHBV/bers, processing time and temperature of all the
fabricated composites are listed in Table 1.
3. Testing and characterization
3.1. Mechanical properties
Tensile and exural properties of specimens were mea-
sured using a United Calibration Corp. SFM 20 testing
machine as per ASTM D638 and ASTM D790 standards,
respectively. Tensile testing was carried out at a rate of
2.54 mm/min with a pre-load of 0.023 kg. The system con-
trols and data analysis were performed using the supplied
Datum software. The notched Izod impact testing was car-
ried out on a Testing Machine Inc., (TMI) model 43-02-01
as per ASTM D256 with a pendulum of impact energy
1.35 J. The numbers of samples tested in each case were ve
for all the respective compositions of composites.
3.2. Thermo-mechanical properties
3.2.1. Dynamic mechanical analysis (DMA)
The storage modulus, loss modulus and tan d of PHBV
and its natural ber composites were measured using a TA
Instruments model Q800 DMA. The specimens were
 S. Singh et al. / Composites: Part A 39 (2008) 875886
heated from 30 to 125 C at a heating rate of 2 C/min.
The single cantilever mode was used to test the specimens
at amplitude of 15 lm and frequency of 1 Hz.
3.2.2. Heat deection temperature (HDT)
The DMA was used to determine the HDT as per
ASTM D648. Rectangular bars 2  12  58 mm in three-
point bending mode with an applying load of 66 psi
(455 KPa) were used for testing. Samples were heated at
the rate of 2 C/min from room temperature to the temper-
ature required to produce 0.195% strain.
3.3. Thermal properties
3.3.1. Thermo-gravimetric analysis (TGA)
TGA was tested using TGA 2950 instrument with a
heating rate of 20 C/min from room temperature to
600 C, the temperature of complete thermal degradation.
All TGA runs were performed under a nitrogen purge.
3.3.2. Dierential scanning calorimetry (DSC)
DSC of the samples was performed using a model Q100
DSC, TA Instruments. The heat/cool method was adopted
for all specimens testing. Specimens were heated to 190 C
at a ramp rate of 10 C/min and cooled with the same
ramp. The data for melting points and heat of crystalliza-
tion were collected during the heating cycle. The crystalli-
zation temperature and heat of fusion were obtained
from the cooling cycle.
Data analyses of DMA, HDT, TGA and DSC results
were performed using the supplied software Universal
Analysis and the numbers of specimens used for testing
each composition were three.
4. FTIR analysis
A PerkinElmer system 2000 FTIR Spectrometer was
used with an attenuated total reectance (ATR) accessory
to gather the IR spectra of neat PHBV and its natural ber
composites.
4.1. Morphological studies
Morphological studies were conducted using a JEOL
scanning electron microscope (model JSM-6400). Speci-
mens were sputter-coated with gold to a thickness of
10 nm before surface characterization in order to prevent
charging. The SEM was equipped with a lanthanum hexab-
oride (LaB6) crystal as electron emitter source. An acceler-
ating voltage of 13 KV was used to collect the SEM images.
4.2. Statistical analysis
Statistical analysis of tensile and exural modulus
results was carried out by performing two-way ANOVA
on the PHBV and its composites with wood and bamboo
ber at 30 and 40 wt.% using MINITAB software.
877
5. Results and discussions
5.1. Tensile properties
Fig. 1 shows the Youngs modulus of PHBV and its
composites with bamboo ber at 30 and 40 wt.% loadings.
The modulus of the PHBV composites increased steadily
with ber loading. At 30 wt.% ber loading it increased
by 67% and at 40 wt.% the increase was 175%. Tensile
properties of short ber composites depend on ber length,
distribution, orientation and their interfacial bond strength
with matrix. The ber orientations have been assumed to
be random; as observed visually and composites behave
isotropically. The increased modulus of the composites
can be ascribed to uniform dispersion of bers in the
PHBV matrix; which leads to the even distribution and
transfer of stress from matrix to ber. The uniform disper-
sion of bers was physically visible in specimens. The ten-
sile strength of PHBV decreased with the addition of
bamboo ber. This can be attributed to the lack of su-
cient interfacial interaction between the ber and matrix.
The presence of polar group on the lignocellulosic natural
bers and the abundant moderate polar groups in the
chemical makeup of PHBV lead to an expectation of some
interfacial interaction between the two; although from the
FTIR analysis, no chemical or hydrogen bond interactions
between the moderately polar groups of the two compo-
nents were observed. Reader may refer Fig. 7 for this pur-
pose. Therefore, the only possible interactions responsible
for the load transfer from matrix to ber were frictional
and Van der Waal forces. To increase the interfacial bond-
ing between the ber and matrix use of compatiblizer lies
under the scope of future work.
5.2. Christensens model
The theoretical Youngs modulus of the composites has
been estimated using Christensens model of equations
[16,17]. It provides a solution for the elastic properties of
composites with randomly oriented short bers. The fol-
lowing expression predicts the elastic modulus of compos-
ites below and above the critical length (lc) of the ber.
3.5 25
Tensile Strength (MPa)
Tensile Modulus (GPa)
3
20
2.5
2 15
1.5 10
1
5
0.5
0 0
Neat PHBV 30/70 (BF/PHBV) 40/60 (BF/PHBV)
Fiber/PHBV (wt%)
Fig. 1. Tensile properties of bamboo berPHBV composites at 30 and
40 wt.% ber, (j) tensile modulus, () tensile strength.
878 S. Singh et al. / Composites: Part A 39 (2008) 875886
 
1 lc crystallinity, which had been noted from the FTIR and
E c  vf E f 1  vm E m at l > lc
6 2l DSC analysis.
1 l
E c  vf E f vm E m at l < lc 5.3. Flexural properties
6 2lc
Flexural modulus and strength of PHBV and its bamboo
Ec denotes the modulus of the composite, Ef and Em that of ber composites with dierent ber loadings are shown in
ber and matrix (in this case bamboo ber and PHBV), vf Fig. 3. In general, the modulus increased with incorpora-
and vm are the volume fractions of ber and matrix, respec- tion of bers. At 30 wt.% of ber content the PHBV mod-
tively, l is the ber length and lc the critical ber length ulus improved by 100% and the increase was greater at
which is given by an expression of the Rosen model. For 40 wt.% ber with the modulus jump of 154%. The exural
a bamboo ber lc < l. strength of PHBV remained unchanged upon the addition
" 1=2 # of bers. The improved exural modulus and retained
1=2 
lc 1  vf Ef strength of the composites validates the reasonable amount
 1:15 1=2
df vf Gm of compatibility and interfacial interaction of cellulose ber
to the natural polyester PHBV. Fiber orientation in the
d f is the diameter of the fiber
matrix plays a vital role in determining the exural proper-
ties. Unlike tensile testing (where induced tensile stress
Gm has been estimated from a general relation plays crucial role) the exural testing incorporates both
E
G m is Poissons ratio compression and tensile stresses on the ber, matrix and
21 m their interface across the specimen cross section, therefore
Fig. 2 depicts the theoretical and experimentally deter-
the strength response to the stress during exural testing
mined values of the modulus. Theoretical values of ligno-
would be higher than during tensile testing. The tensile
cellulosic natural bers (bamboo) and the PHBV matrix
strength is chiey govern by the dewetting of matrix from
have been approximated to be 25 GPa [18] and 1.02 GPa,
the ber under tensile load and also is more sensitive to
respectively. Volume fraction of the ber and the matrix
the defects originated in the interfaces of matrix and ber.
at 30 and 40 wt.% were calculated using general equations
mf
qf
vf mf and vm = (1  vf) assuming density of natural 5.4. Dynamic mechanical analysis (DMA)
mqmm qf
ber (qf) and matrix (qm) to be 0.84 gm cm3 [19] for
The viscoelastic characteristics of the materials were
bamboo ber and 1.25 gm cm3 for PHBV. The modulus
measured using DMA. It is a complex quantity dened
of bamboo berPHBV composites as predicted by
with elastic (storage) and viscous (loss) components;
Christensens model are in close approximation to the
denoted E0 and E00 , respectively. In DMA, the eect of tem-
experimental results. This was due to the uniform and nar-
perature on the viscoelastic properties of the composites
row size distribution of bamboo bers, although not quan-
was studied. This gives more distinct insight to the interac-
tied but observed visually. The dierence between the
tions at the molecular level of the materials.
theoretical and experimental data can be due to the
Fig. 4 depicts the change in the E0 of PHBV and its com-
agglomeration of the llers at high volume fractions which
posites with temperature. Like a generic tendency of ther-
results in the higher modulus than the expected. The
moplastic composites, E0 of PHBV and its composites
agglomeration was visible in SEM micro photographs.
decreases with rising temperature. E0 of PHBV increased
Although the composites have been assumed to be isotro-
with the ber loading up to 40 wt.%. The dierence between
pic; anisotropy of the bers cannot be ruled out. The model
the modulus values of bamboo ber composites at 30 and
does not account for the eect of ller and processing con-
40 wt.% was not signicant. E0 of bamboo ber composites
ditions on the matrix (PHBV), e.g., change in the degree of

3.5
Flexural Strength (MPa)
Flexural Modulus (GPa)

3.5
Elastic modulus (GPa)

3 30
3
2.5
2.5
2 20
2
1.5
1.5
1 1 10

0.5 0.5

0 0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 Neat PHBV 30/70 (BF/PHBV) 40/60 (BF/PHBV)
Volume fraction of fiber Fiber/PHBV (wt%)

Fig. 2. Experimental and theoretical data plots. h PHBVbamboo ber Fig. 3. Flexural properties of bamboo ber/PHBV composites having 30
composite, () Theoretical PHBVbamboo ber composite. and 40 wt.% ber, (j) exural modulus, () exural strength.

7000
6000
Storage Modulus (MPa)
5000 (b)
4000
3000
(a)
2000
1000
0 ites (Fig. 6) occur in the glass transition temperature region.
-25 0 25 50 75
Temperature ( C)
Fig. 4. Storage modulus versus temperature of PHBV with varying
loading level of ber (a) PHBV, (b) PHBV/bamboo ber (70:30) wt.% and
(c) PHBV/bamboo ber (60:40) wt.%.
improved by 157% and 173% at 30 and 40 wt.% ber con-
tent at 25 C, respectively. The increase in E0 of the ber
lled PHBV with respect to neat PHBV was greater below
Tg than above it. This can be attributed to the temperature
sensitivity of the natural ber. The slope of the E0 versus
temperature curve around the glass transition (Tg) temper-
ature (1 C) of PHBV increases with the ber loading.
This gave an indication of increased brittleness of these
composites. A small bump in the curve of PHBV at 52 C
is indicative of the cold crystallization region. During the
heating cycle as the specimen enters the cold crystallization
region, the increasing crystallinity leads to decrease in the
reduction rate of E0 as a result a small bump emerges in this
region. The cold crystallization temperature (Tcc) decreases
with ber loading. The Tcc decreases to 46 C in case of
bamboo ber composites. This could be due to very low
diameter or high aspect ratio of the bamboo ber versus
the wood bers which gave it a better opportunity to act
as nucleating agent in PHBV matrix. Among the dynamic
mechanical properties, dampening is most sensitive to the
structural heterogeneity, morphological and transitions
changes of system, and bermatrix interfacial interactions.
Fig. 5 shows the dampening of the PHBV and its ber com-
posites. Tan d expressed as the ratio of E00 to E0 gives
account of the energy dissipated as heat during the dynamic
testing. Tan d peak occurs approximately 10 to 20 C above
the Tg of the PHBV and its ber composites. Increased sti-
0.1
(a)
0.08
Tan delta
0.06 (b)
(c)
0.04
0.02
0 0
-25 0 25
Temperature (C)
Fig. 5. Tan d versus temperature of PHBV with varying loading level of
ber (a) PHBV, (b) PHBV/bamboo ber (70:30) and (c) PHBV/bamboo
ber (60:40).
S. Singh et al. / Composites: Part A 39 (2008) 875886 879
ness of the PHBV with the ber loading decreases the tan d
peak value thus reecting the reduced energy losses in the
interfacial regions. Low dampening of the composites in
the transition region is attributed to the neness of the bam-
(c) boo bers which allows their uniform distribution and crys-
tallinity eect on the matrix. The widening of the tan d curve
also indicates the increased heterogeneity of the specimens
which was observed from the SEM photomicrographs.
The maxima of the loss modulus of PHBV and its compos-
100 125 150 The enhanced response of molecular chains to the tempera-
ture in the amorphous region increases the viscous charac-
teristic of the material in the glass transition domain. The
ber incorporation into the PHBV slightly increases its Tg
towards higher temperature, thus delaying the increasing
viscosity component of the composites. This can be due to
the restriction imposed by the bers on the movement of
the expanding molecular chains in the amorphous region.
5.5. FTIR analysis
In the PHBV IR spectra, Fig. 7a, the band at 1720 cm1
was the C@O stretch of the ester group present in the
molecular chain of highly ordered crystalline structure
[20]. A small hump at 1740 cm1 was of the C@O stretch
in the amorphous region of semicrystalline PHBV. Charac-
teristic peaks of symmetric COC stretching vibration
were present from 800 cm1 to 975 cm1 and the antisym-
metric COCstretching between 1060 and 1150 cm1
[21]. The band around 1378 cm1 corresponds to the sym-
metrical wagging of CH3 groups and at 1453 cm1 was due
to asymmetric deformation of methylene groups, the inten-
sity of both the peaks is independent of crystallinity or in
other words insensitive to crystallinity. The absorption
bands at 1720, 1276, 1228, and 980 cm1 arose from the
crystalline regions (designated with arrows) and bands at
1740, 1453, 1176 cm1 from the amorphous regions (desig-
nated with two face arrows) of PHBV. Peak at 1228 cm1
was purely of the conformational band of crystalline helical
molecular chains and therefore are absent in amorphous
region [22]. The crystallinity index (CI) gives the relative
300
250
Loss Modulus (MPa)
200
150
(c)
100
(b)
50
(a)
50 -25 0 25 50 75 100 125 150
Temperature (C)
Fig. 6. Loss modulus versus temperature of PHBV with varying loading
level of ber (a) PHBV, (b) PHBV/bamboo ber (70:30) wt.% and (c)
PHBV/bamboo ber (60:40) wt.%.
880 S. Singh et al. / Composites: Part A 39 (2008) 875886

(b)

(a)

1750 1550 1350 1150 950 750

Fig. 7. FTIR spectra of (a) PHBV and (b) PHBV/bamboo ber (70:30) composite.

measure of degree of crystallinity in PHBV. It is given as 120

I 1
CI at 1378cm (ratio of intensity of the band insensitive 100
I at 1176cm1
Percentage weight loss (%)

to crystallization to the band sensitive to crystallization),

80
where I is intensity of the peak at a given wave number. (c)

Fig. 7b, (bamboo and PHBV composites) depicts the IR 60

spectrum of bers and PHBV composites. On the compar-
ison of the intensity of peaks at 1228 cm1 and 1176 cm1 40
(b)
of PHBV and its ber composites, it is observed that these
vary inversely to one another with the changing degree of 20 (a)

crystallization. The CI of PHBV 0.66 increased to 0.74 with

0
the addition of 30 wt.% bamboo ber indicating the 0 100 200 300 400 500
increase in the crystallinity of PHBV when the bamboo Temperature (C)
ber was induced. CI value of similar order was obtained
Fig. 8. Percentage loss of (a) PHBV, (b) PHBV/bamboo ber (70:30) and
even though when a 40 wt.% bamboo ber was added to (c) bamboo ber.
the PHBV.

5.6. Thermo-gravimetric analysis (TGA)
The TGA curves in Fig. 8 depict the weight loss of
bers, PHBV and its composites with temperature. Bam-
boo bers began to thermally degrade at 250 C, reaching
the maximum at 380 C and entirely degrading at around
500 C. The PHBV lost weight gradually with the increas-
ing temperature and its thermal degradation occurred max-
imum at 316 C while insetting at 250 C. Its thermal
instability above 250 C has earlier been reported stating
that the degradation process involves chain scission and
hydrolysis which leads to reduction in molecular weight
and formation of crotonic acid [23] and the presence of
moisture further promotes hydrolytic degradation. The
thermal degradation of biocomposites appeared to be
cumulative phenomena of thermal degradation of matrix
(PHBV) and ber alone. Degradation of composites initi-
ated around 250 C in case of both, ber and PHBV and
ceased around 400 C. The TGA of composites also
showed the bers to be nal constituent to undergo degra-
dation in composites. The composite degradation maxima
occurred at temperature a little higher than that of PHBV.
Fig. 9 illustrates the derivative weight loss of PHBV, bers
and their composites at 30 wt.% ber. Maximum degrada-
tion temperature peak of the composite shifted to the
higher temperature from that of PHBV indicating it to be
a slightly more thermally stable than PHBV. Thus the pres-
ence of bamboo ber (a lignocellulosic ber) did not have
any degradation eect on PHBV.
5.7. Heat distortion temperature (HDT)
The HDT is an upper use temperature limit of a mate-
rial, therefore, the eect of ber type and loading in com-
posites has a considerable importance in their structural
applications. The HDT of semicrystalline polymers like
 S. Singh et al. / Composites: Part A 39 (2008) 875886 881

4.5 40

Izod Impact Strength (J/mt)

4 35
Derivative wt loss (mg/(C )

(a) 30
3.5
25
3
20
2.5 15
2 10

1.5 5
(b) 0
1
Neat PHBV 30/70 (BF/PHBV) 40/60 (BF/PHBV)
0.5 (c) Fiber/PHBV (wt%)
0
0 100 200 300 400 500 600
Fig. 11. Notch impact strength of PHBV/bamboo ber composites at 30
Temperature (C)
and 40 wt.% loading levels.

Fig. 9. Derivative weight loss of (a) PHBV, (b) PHBV/bamboo ber respect to ber to matrix contact, which did not allow
(70:30) and (c) bamboo ber. the sucient ber matrix interfacial interactions to
develop. As seen from the micrograph discussed in the fol-
PHBV depends on thermal history, crystallite sizes and dis- lowing section majority of the fracture occurred through
tribution along with the presence of impurities and other the ber rather than through the matrix or the interface
nucleating agents which aects their crystallinity [24]. of matrix and ber.
Fig. 10 has a description of the HDT values of PHBV
and its composites with bamboo bers at 30 and 40 wt.% 5.9. Morphology
ber content. The HDT of PHBV was enhanced by addi-
tion of bers at 30 and 40 wt.% from 114 to 120 C and Fig. 12a and b shows the micro photographs character-
123 C, respectively. The major contributing factor to the ized under SEM at dierent magnications of the impact
improved HDT with higher degree of ber content was fractured surfaces at 30 wt.% of bamboo ber composites.
the ber reinforcement which had better HDT than the
matrix. The increased storage modulus of composites at
higher temperatures also contributed to the increased
HDT. Similar types of observations have been earlier
reported [11,15].

5.8. Impact strength

Impact strength depends on various factors like ber

matrix adhesion, toughness of the matrix and ber, defects
in the packing of ber/matrix, crystalline morphology, etc.
[25]. The notch Izod impact strength test measures the
energy to propagate an existing crack. Fig. 11 depicts the
notched Izod impact strength of bamboo ber composites
with PHBV. The impact strength of PHBV (34.3 2.7 J/
mt) decreased with the addition of ber but there was no
signicant reduction when the ber content was increased
from 30 to 40 wt.%. The higher ber volume in matrix gave
manifold ber to ber contact (ber clustering) with
130

120
HDT (C)

110

100
Neat PHBV 30/70 (BF/PHBV) 40/60 (BF/PHBV)
Fiber/PHBV (wt%)

Fig. 10. HDT of bamboo ber/PHBV composites having 30 and 40 wt.% Fig. 12. SEM photomicrographs of impact fractured samples of PHBV/
ber. bamboo (70:30) composite: (a) 90, 200 lm and (b) 450, 50 lm.
882 S. Singh et al. / Composites: Part A 39 (2008) 875886

and separation during impact. Fibermatrix adhesion

remained below the critical adhesion limit, due to which,
the impact energy decreased [26].

6. Analogy of bamboo berPHBV and wood berPHBV

biocomposites

There was not any signicant dierence among the ten-

Fig. 13. SEM photomicrographs of impact fractured samples of PHBV/
bamboo (60:40) composite: (a) 90, 200 lm and (b) 500, 50 lm.
The failure modes observed during the fracture were ber
pull-out and ber breakage. Due to the higher volume frac-
tion of bamboo ber the majority of the surface could be
seen dispersed with the bers along with the clustering. A
number of voids were also visible on the fractured surface
that had contributed to the low impact and tensile strength
with the ber loading. Fiber separation in the clusters con-
tributes much lesser to the energy than the ber pull-outs
from the matrix in the total fractured energy during the
impact. This was consistent with the observation of low
impact strength of the bambooPHBV composites,
(Fig. 13). At 40 wt.% ber content ber clustering domi-
nated over the ber matrix adhesion thus giving insucient
adherence of ber to matrix which lead easy ber pull-out
sile properties of bamboo and wood ber based composites
at the same loading level even though the aspect ratios of
the two bers were dierent (length and diameter of the
wood bers were between 1600 and 1650 lm and between
300 and 400 lm [15] and that of bamboo ber were
500 lm and 10100 lm, respectively). Table 2 lists vari-
ous mechanical properties of the two biocomposites at 30
and 40 wt.% ber. The exural strength of both types of
biocomposites remained somewhat unaltered with the
addition of bers up to 40 wt.% and the modulus in case
of wood ber composites was greater than the bamboo
ber composites by 14%. This may be due to the higher
order stiness of maple wood ber versus bamboo ber.
Furthermore, the eect on tensile and exural modulus of
ber type (wood and bamboo) and loading level (30 and
40 wt.%) of natural bers in PHBV were statistically ana-
lyzed by performing a two-way ANOVA. Table 3 repre-
sents the F, F critic and P values of the eect of class of
ber, their loading level and an interaction of the type
and loading levels on the afore mentioned characteristics.
ANOVA was performed at a signicance level (a) of 0.05.
From the listed values in the Table 3, it can be inferred that
the type of ber does not have any signicant eect on the
elastic and exural modulus of composites whereas the
ber loading at 30 and 40 wt.% have signicant eect on
both moduli. The interaction eect of both parameters on
tensile and exural modulus varies. For the tensile modulus
the interaction eect of ber class and loading was signi-
cant but for the exural modulus it was insignicant. This
is consistent with the experimental results because the ten-
sile properties are more sensitive to the defects (stress con-
centration) originated during processing or fabrication of
the specimens.
The HDT of the wood ber biocomposites was greater
than that of the bamboo ber biocomposites by 15 C at
both the loadings levels, whereas the increase from 30 to

Table 2
Description of bamboo and wooda ber based PHBV biocomposites properties
wt.% of ber in PHBV Type of Tensile properties Flexural properties Impact strength (J/mt) HDT
matrix ber (C)
Strength Modulus Strength Modulus
(MPa) (GPa) (MPa) (GPa)
0 (neat PHBV) 21.4 1.02 30.27 1.28 34.3 114
30 Bamboo 18.9 1.71 30.6 2.56 24.28 120
30 Wood 18.08 1.94 31.09 2.72 30.0 135
40 Bamboo 16.74 2.8 29.59 3.26 23.75 123
40 Wood 16.75 2.73 29.97 3.44 24.4 137
a
The data of the wood berPHBV composites has been adapted from the earlier publication of the authors for the purpose of comparison [15].
 S. Singh et al. / Composites: Part A 39 (2008) 875886 883

Table 3
Description of P and F values
Tensile modulus Flexural modulus
P- F F P- F F
value critic value critic
Fiber type 0.247 1.41 4.259 0.058 3.95 4.259
Fiber loading 0 235.42 3.402 0 474.23 3.402
Interaction between 0.031 4.03 3.402 0.386 0.99 3.402
ber type and ber
loading

40 wt.% was a marginal of 2 C in both type of natural

ber biocomposites. The dierence between the HDT of
bamboo and wood ber composites can be attributed to
more uniform distribution of low density and small ber
size of bamboo ber than wood ber which increased the
volume fraction of bamboo ber with respect to matrix
in the composite. This did not allow the sucient matrix
packing and crystallite size growth as it happened in case
of wood ber based composites. This was supported by
the FTIR and DSC analysis depicting the increased crystal-
linity of wood ber based composite versus the bamboo
ber based PHBV composite.
From the comparative TGA of both kinds of compos-
ites, the thermal stability of PHBVbamboo ber compos-
ites was a little more than that of PHBV-wood ber
composites [15]. The reduced thermal stability of PHBV
in presence of wood ber has earlier been reported in
publication [15]. This can be due to the lesser amount
of lignin in bamboo ber than maple wood ber (Table
5). The eect of delignication on the mechanical proper-
ties of PP based WPC has been reported by Alinaghi
et al. [27].
The impact strength of PHBV composites with wood
ber was observed to be higher than that with bamboo
ber. This can be reasoned due to the higher volume frac-
tion of bamboo ber than wood ber at same wt.%. As
observed from FTIR and DSC results, low crystallinity in
the bamboo ber biocomposites may had resulted the
lower impact strength than the wood bers biocomposites.
Wood ber surface gave a nucleation growth eect thus
giving a better embedment of ber in the matrix.
6.1. Dierential scanning calorimetry analysis of bamboo
berPHBV and wood berPHBV biocomposites
Non-isothermal DSC studies of PHBV and its ber
composites was carried out in order to have an understand-
ing of the eect of bers on the nucleation and crystallinity
of PHBV. Fig. 15 shows the thermograms of PHBV and its
30 wt.% wood and bamboo ber composites. The details
are given in Table 4 where Tm is the melting temperature
and Hc is the cold crystallization enthalpy. A bimodal
endotherm of PHBV during the heating cycle is due to
the crystallization phenomena of the heterogeneous crystal
morphology formed during processing. Renstad et al. [28]
Fig. 14. SEM Photomicrographs at high magnication (a) PHBV/wood
ber composite 1500, 20 lm and (b) PHBVbamboo ber composite,
2300, 10 lm.
have carried out a detailed study of this. Embedding the
bers in PHBV does not vary its melting point which
remains at 154 C for both kinds of ber composites. The
enthalpy of melting or crystallization (Hf) in both the cases
of bers composites remains same, thus depicting the
degree of crystallization was unaected by the type of ber
in the PHBV. Mergence of the second peak with the pri-
mary peak of PHBV was an observable phenomenon in
30 wt.% bamboo ber composite which was absent in
30 wt.% wood ber composite. The ne bamboo ber inter-
feres in the crystal morphology of PHBV during the pro-
cessing as a result the nucleation growth of PHBV
became more homogenized in presence of ne and uni-
formly blended bamboo bers in the matrix. Spherulite
homogenization of PBS in presence bamboo ber had been
reported by Kori et al. [29].
Higher volume fraction of bamboo than wood ber at
same wt.% may have increased the nucleating density of
the matrix but no appreciable dierence between the net
crystallinity of both types of composites was observed. It
is to be noted that no such eect was noticed at 40 wt.%.
The melt crystallization temperature (Tc) of PHBV was
observed to be 101 C. The addition of 30 wt.% wood ber
increased this temperature to 104 C, whereas the bamboo
ber decreased it to 99 C; indicating an increase in the rate
884 S. Singh et al. / Composites: Part A 39 (2008) 875886

Heat flow Exo up

(a)

(b)

(c)

60 80 100 120 140 160

Temperature (C )

0 50 100 150 200

(a)
Heat flow down

(c)

(b)

Fig. 15. Crystallization and melting curves of (a) PHBV, (b) PHBV/bamboo ber (70:30) composite and (c) PHBV/wood ber (70:30) composite.

Table 4
boo ber to the nucleus thus decreasing the crystallinity
Detailed information obtained from dierential scanning calorimetry of
PHBV and its composites rate in case of bamboo ber composite. Similarity in the
melt crystallization peaks of PHBV and its brous compos-
Tm (C) Hf (J/g) Tc (C) Hc (J/g)
ites during cooling provided the evidence of having any
PHBV 154 57.25 101.4 48.6
divergence in their crystalline morphologies.
PHBVwood ber (70:30) 153.3 63.28 104.3 35.47
PHBVbamboo ber (70:30) 152.9 62.29 99.1 30.52 FTIR of the bamboo and wood ber were carried out to
PHBV bamboo ber (60:40) 152.7 59.44 97.2 28.92 determine an estimate of their contents and composition.
Fig. 16 shows FTIR spectra of wood and bamboo bers
(described peaks marked with arrows). The region around
of crystallization in the presence of wood ber and decrease 3400 cm1 depicted hydrogen bonded OH stretching and
in the case of bamboo ber. This may be due to slower dif- at 2900 cm1 CH stretching. The peak at 1740 cm1 was
fusion of molecular chains though the well dispersed bam- the C@O stretch in hemicelluloses present in both lignocel-

Wood fiber
A
b
s
o
r
b
Bamboo fiber
a
n
c
e

3700 3200 2700 2200 1700 1200 700

Waven umber (cm-1)

Fig. 16. FTIR spectra of wood and bamboo ber.

 S. Singh et al. / Composites: Part A 39 (2008) 875886
Table 5
Lignin and cellulose content of wood and bamboo bera
Fiber type Lignin (%) Cellulose (%)
Wood 3849 2330
Bamboo 2131 2643
a
Ref. [31].
lulosic bers. 1505 and 1600 cm1 are characteristic peaks
of lignin which arises due to the aromatic skeletal vibration
of benzene ring in lignin. The peak at 1600 cm1 has con-
tributions from the conjugated C@O group along with aro-
matic skeletal stretch [30]. The ratio of the two (1505
1600 cm1) gives some qualitative aspects of two dierent
types of lignins present in dierent kinds of natural bers.
Similar spectrums of wood and bamboo ber suggested the
presence of same proportion of cellulose, hemicellulose and
lignin in both types of natural bers, as supported in the lit-
erature [30]. Composition of lignin and cellulose in both
bers are listed in Table 5 [31]. Wood ber composites
had a higher CI values (0.95) than bamboo ber compos-
ites at 30 wt.% ber content indicating greater crystallinity
of PHBV in presence of wood ber than bamboo Similar
results were noticed at 40 wt.% ber content in both type
of composites [15].
Fig. 14 are the photomicrographs at higher magnica-
tion of composites depicting the bers embedded in the
matrix and its wettability on the ber surface indicating
the similar interfacial interactions existing between the
two dierent lignocellulosic ber and PHBV matrix.
7. Conclusions
Biocomposites from a bamboo ber and poly(hydroxy-
butyrate-co-hydroxyvalerate) (PHBV) were fabricated
using extrusion followed by injection molding. Tensile
and exural modulus of this biocomposite increased with
ber loading. The storage modulus also improved with
ber addition and no appreciable dierence among the
two compositions (30 and 40 wt.% ber) was noticed.
Comparative analysis of two dierent type of lignocellu-
losic ber (wood and bamboo) at 30 and 40 wt.% ber
loaded biocomposites were studied with respect to their
mechanical, thermo-mechanical, thermal, FTIR and mor-
phological aspects. Both types of lignocellulosic bers
embedded in the PHBV matrix gave an appreciable rise
in tensile and exural moduli. Statistically, there was no
eect of the ber type on mechanical properties of compos-
ites. A dierence between the net crystallinity of both types
of composites with dierent bers was observed in dieren-
tial scanning calorimetry (DSC) and infrared spectroscopy
(FTIR). Notch impact strength of PHBV decreased with
the ber addition and the reduction was greater in case
of bamboo ber composites. No eect of the ber type
on the thermal stability of composites was observed, but
incorporation of wood ber gave a higher heat deection
temperature (HDT) versus the bamboo ber. Scanning
885
electron microscopy (SEM) photomicrographs demon-
strated the existence of similar kinds of interfacial interac-
tion between the two dierent types of bers and PHBV.
Acknowledgements
Authors are grateful for the nancial support provided
by the USDA-CSREES under the McIntire-Stennis pro-
gram of the Michigan Agricultural Experiment Station,
Michigan State University. We acknowledge American
wood bers (Schoeld, WI, USA) and Biomer (Germany)
for supplying maple wood ber and PHBV, respectively.
References
[1] Karus M, Ortmann S, Vogt D. Use of natural bers in composites in
the German automotive production 1996 till 2003 (200410); <http://
www.nachwachsende-rohstoe.info> (date accessed: 12/4/06).
[2] Ferrer FM, Vilaplan F, Greus AR, Borras AB, Box CS. Flour rice
husk as ller in block copolymer polypropylene: eect of dierent
coupling agents. J Appl Polym Sci 2006;99:182331.
[3] Huda MS, Mohanty AK, Drzal LT, Shut E, Misra M. Green
composites from recycled cellulose and poly(lactic acid): physico-
mechanical and morphological properties evaluation. J Mater Sci
2005;40:42219.
[4] Rana AK, Mitra BC, Banerjee AN. Short jute ber-reinforced
polypropylene composites: dynamic mechanical study. J Appl Polym
Sci 1999;71:5319.
[5] Lu JZ, Negulescu II, Wu Q. Maleated wood-ber/high-density-
polyethylene composites: coupling mechanisms and interfacial char-
acterization. Compos Interf 2005;12(12):12540.
[6] Shu-Kai Yeh, Casuccio Mary E, Al-Mulla A. Polypropylene-wood
plastic composites reinforced with dierent amounts of wood and
organoclay. In: Annual technical conference society of plastics
engineers, vol. 64; 2006. p. 20913.
[7] Khalil HPSA, Shahnaz SBS. Recycle polypropylene (RPP) wood
saw dust (WSD) composites Part 1: The eect of dierent ller size
and ller loading on mechanical and water absorption properties. J
Reinf Plast Comp 2006;25(12):1291.
[8] Luzier WD. Materials derived from biomass/biomaterials. Proc Natl
Acad Sci USA 1992;89:83942 [Colloquium paper].
[9] Williams SF, Martin DP, Horowitz DM, Peoples OP. PHA appli-
cations: addressing the price performance issue: I. Tissue engineering.
Int J Biol Macromol 1999;25:11121.
[10] Dufresne A, Dupeyre D, Paillet M. Lignocellulosic our-reinforced
poly(hydroxybutyrate-co-valerate) composites. J Appl Polym Sci
2003;87:130215.
[11] Bhardwaj R, Mohanty AK, Drzal LT, Pourboghrat F, Misra M.
Renewable resource based green composites from recycled cellulose
ber and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) bioplastic.
Biomacromolecules 2006;6(7):204451.
[12] Fernandes EG, Pietrini M, Chiellini E. Bio-based polymeric com-
posites comprising wood our as ller. Biomacromolecules
2004;5(4):12005.
[13] Kitagawa K, Ishiaku SU, Mizoguchi M, Hamada H. In: Mohanty
AK, Misra M, Drzal LT, editors. Natural bers, biopolymers and
biocomposites. Taylor and Francis; 2006.
[14] Yongli M, Xiaoya C, Qipeng G. Bamboo ber-reinforced polypro-
pylene composites: crystallization and interfacial morphology. J Appl
Polym Sci 1997;64:126773.
[15] Singh S, Mohanty AK. Wood ber-reinforced bacterial bioplastic
composites: fabrication and performance evaluation. Compos Sci
Technol 2007;67(9):175363.
[16] Berlin AA, Volfson SA, Enikolopian NS, Negmatov SS. Principles of
polymer composites. Berlin, Heidelberg: Springer-Verlag; 1986.
886 S. Singh et al. / Composites: Part A 39 (2008) 875886
[17] Christensen R. Mechanics of composite material. New York: John
Wiley and Sons; 1979.
[18] Deshpande AP, Rao MB, Rao CL. Extraction of bamboo bers and
their use as reinforcement in polymeric composites. J Appl Polym Sci
2000;76:8392.
[19] Rajulu AV, Chary KN, Reddy GR, Meng YZ. Void content, density
and weight reduction studies on short bamboo ber-epoxy compos-
ites. J Reinf Plast Comp 2004;23:127.
[20] Padermshoke A, Katsumoto Y, Sato H, Ekgasit S, Noda I, Ozaki Y.
Surface melting and crystallization behavior of polyhydroxyalkano-
ates studied by attenuated total reection infrared spectroscopy.
Polymer 2004;45:654754.
[21] Suthar V, Pratap A, Rawal H. Studies on poly(hydroxy alkanoates)/
(ethylcellulose) blends. Bull Mater Sci 2000;23(3):2159.
[22] Hong SG, Chen WM. The attenuated total reection infrared
analysis of surface crystallinity of polyhydroxyalkanoates. e-Poly-
mers 2006; 024: (<http://www.epolymers.org>, ISSN 1618-7229).
[23] Gatenholm P, Mathiasson A. Biodegradable natural composites. II.
Synergistic eects of processing cellulose with PHB. J Appl Polym Sci
1994;51:12317.
[24] Takemori MT. Towards an understanding of the heat distortion
temperature of thermoplastics. Polym Eng Sci 1979;19(15):11049.
[25] Kamdem P, Jiang HC, Freed JW, Matuana ML. Properties of wood
plastic composites made of recycled HDPE and wood our from
CCA-treated wood removed from service. Compos Part A
2004;35:34755.
[26] Agarwal BD, Broutman LJ. Analysis and performance of ber
composites. New York: John Wiley and Sons; 1980.
[27] Alinaghi K, Saleh N, Ismaeil G, Mehdi T, Ghanbar E. Eect of the
delignication of wood bers on the mechanical properties of wood
ber-polypropylene composites. J Appl Polym Sci 2006;102:475963.
[28] Renstad R, Karlsson S, Albertsson AC, Werner PE, Westdahl M.
Inuence of processing parameters on the mass crystallinity of
poly(3-hydroxybutyrate-co-3-hydroxyvalerate). Polym Int 1997;43:
2019.
[29] Kori Y, Kitagawa K, Hamada H. Crystallization behavior and
viscoelasticity of bamboo ber composites. J Appl Polym Sci
2005;98:60312.
[30] Pandey KK. A study of chemical structure of soft and hardwood and
wood polymers by FTIR spectroscopy. J Appl Polym Sci
1999;71:196975.
[31] <http://www.PaperListeningStudy.org> (date accessed 03/03/07).